0044 Pleiotropic Effects of APOA2 Variants on a Sleep-Based Adult Epidemiological Cohort Implicated in Sleep Breathing and Memory Phenotypes

Abstract Introduction APOA2 gene encodes a subclass of apolipoproteins implicated in the regulation of risk factors for cardiovascular disease and obesity. Functional studies in animal models have demonstrated that APOA2 genetic variants are involved in breathing instability and therefore might confer risk for sleep apnea. Methods Phenome Wide Association study (PheWAS) approach was applied in São Paulo Epidemiologic Sleep Study (EPISONO), an adult epidemiological sample (1,042 individuals, 20-80 years-old) to address effects of APOA2 SNV rs5082. EPISONO sample was subjected to SNP-array genotyping and deep phenotyping, including objective and subjective sleep evaluations, laboratory tests, clinical scales, anthropomorphic measurements and sociodemographic inquiries. An additive genetic model contrasted rs5082 genotypes with 460 traits. Different regression models were applied depended on the nature of the phenotypic variable. Covariates included age, sex, principal components (PCs 1-10), and the Brazilian socioeconomic classification questionnaire. In a second round, body mass index (BMI) was included as an additional covariate. To correct for multiple testing, Bonferroni test was applied and non-parametric tests were used to validate nominally significant associations. Results Fifteen suggestive associations (p< 0.05 in PheWAS) were validated by non-parametric tests. In the biochemistry profile, high-density lipoprotein, and eosinophils/eosinophils count were linked to allelic differences. Of polysomnography measures, respiratory effort-related arousals (RERAs) were linked. Prospective memory total score and memory complaints evaluated by the Prospective and Retrospective Memory Questionnaire were also implicated. The score for risky tobacco use measured by Alcohol, Smoking and Substance Involvement Screening Test was associated with genotype. When adjusting for BMI, an inquiry about overall satisfaction about health conditions remained significant after Bonferroni correction. Conclusion Known APOA2 SNP associations with lipid metabolism (HDL dosage) were replicated in an admixed cohort, despite sample size limitations. A lack of a direct effect on obesity may be explained by dietary heterogeneity (information not collected). Suggestive associations are present concerning variant effect on expression of RERAs, consistent with the deterministic effect of APOA2 expression observed mouse models. Memory associations imply that APOA2 may affect cortical function, consistent with its apolipoprotein role in cholesterol transport. Support (if any) AFIP, FAPESP, CNPq.